Advertisement

Many products and technologies are specialized in the collection of evidence while others' sole purpose is to keep evidence organized. However, even fewer technologies can collect, organize, and provide the analysis tools all in one package. This is the main reason why 3D scanning for forensics is an ever growing and useful application of laser based measurement technologies in fighting crimes and reconstructing events. Like all forensic tools, however, 3D scanning has its strengths and its limitations.

Benefits of 3D Scanning in Forensics
It’s important to distinguish between data and evidence when capturing data with 3D scanners. 3D scanners are indiscriminate in their method of capturing data and can be viewed as "spray painting" the environment with thousands of dot measurements. However, not all data is evidence and unlike using a total station or taking hand measurements, a forensic technician can do a very detailed scan of an entire crime scene without the need to predetermine what areas of the scene contain true evidence. In some scenarios, evidence is quite obvious, but in others, "evidence" can be questionable or misleading. These often missed areas are captured in detail such that the resulting data set allows the crime scene to be reviewed at a later day by a forensic expert, who determines what is indeed evidence.

Aside from scanning speed, 3D scanners are particularly well suited for scanning organic shapes and highly curved surfaces that would otherwise be difficult to measure. In the case of bloodstains or "print" evidence that can span over several surfaces, furniture, and walls, trying to accurately measure evidence by hand can prove to be extremely tedious, inaccurate, and time consuming. Even with the use of a total station, the number of data points required would result in far too much time and effort when compared to the use of a 3D laser scanner.

estimated scanning rates using common equipment

A perfect example of this is in the analysis of bite marks located around the curved surface of an arm or other body part that is not perfectly flat. Photography is the traditional method of documenting bite marks, but there is no depth information contained in a photograph. Photography effectively takes 3D objects in the real world and projects them on a 2D image plane (i.e. the CCD on your digital camera) so that depth information is lost. The true nature of a bite mark can best be measured through the use of a close range scanner where the mark may wrap around the surface of an arm or hand. Thus, having the depth information and proper orientation of the bite mark on the surface of the body can provide the best possible "match" to a suspect.

Photography and casting are effective ways to capture footprints and other impression evidence. However, when the substrate is not a firm soil or hard material, the act of pouring a casting liquid into a print can effectively deteriorate the quality such that details are lost.

Contamination and the ability to physically reach a point for measurement are often issues with hand measurements. Many total stations now have the ability to capture point measurements in "reflectorless" mode, however using a pole and prism is limited to the accessibility of the surface being measured. Also, one would not want to touch a decaying body or fresh bloodstain until proper documentation procedures have been completed. 3D Scanners and photogrammetry measurements are convenient this way since they are non-contact methods that can easily gather measurements from hard to reach places. The general rule of thumb is that as long as the surface of the object is within the line of site of the instrument, then it can be measured.

Of course, being a non-contact form of measurement means that the scanning technician does not need to be physically near or touch any part of a collapsed bridge, decaying body, or chemical spill. Most measurements can be done from a safe distance without the need to be directly at the point of interest.

comparison: photograph, point cloud, and rendered mesh

Very small blood spatter is difficult and time consuming to measure by hand. Documenting the evidence using 3D techniques saves time and provides more accurate results.

Since laser scanners emit a beam of laser light, there is some concern by equipment operators and the general public about direct eye contact with the laser causing permanent visual damage. There are several laser classes, but fortunately, most 3D laser scanners are now classified as "eye safe". Class 1: Inherently safe; no possibility of eye damage. This can be either because of a low output power—in which case eye damage is impossible even after hours of exposure—or due to an enclosure preventing user access to the laser beam during normal operation, such as in CD players or laser printers.1 However, most practitioners would agree that you should not look directly into a laser emitting device just as you would not stare directly at the sun on a sunny day.

A 3D scan of a footprint in sand can give a better chance at a "match" since depth information is available.

More recently, the two most common methods of documenting a crime scene have traditionally been hand measurements and by using a total station to record specific points and features. In both of these methods, the data is strictly a list of numbers on paper or in a digital file. Although many total stations now come with modern data collectors that can give a 3D representation of the points as they are collected, further work is required to turn the data into something visually meaningful by use of a computer aided drafting package or some other specialized crime scene drawing software. In this respect, 3D scanners are different (even from photogrammetry solutions) since many of the newer 3D scanners have on board displays that give real-time visual feedback as the points are being captured.

Both photogrammetry and 3D scanners can automatically collect the color (RGB: Red, Green, Blue) information for each point measured. Traditionally, measurements taken by hand or with a total station are simply "white points" when translated and viewed in a CAD package. Further work is required with traditional methods in order to connect points with lines and draw surfaces between points.

One of the major benefits of modern 3D scanners it that they have become relatively easy to setup and capture data when compared to total stations. 3D scanners do not require backsights or reference points and can be easily picked up and moved to a new location for scanning. Often, reference points/targets are used to determine the scanner's position for each of the scans. Combining all the scans into one large scan is usually automated through software and when there is enough overlap between or when reference targets have been captured. Then, it is possible to accurately "register" scans (i.e. combine separate point clouds together).

3D scanners are special in the sense that in almost all cases, once the scene is scanned, it is immediately available for review and can be analyzed on a laptop computer at the scene. Areas lacking detail can be re-scanned while other areas that did not scan properly can be addressed.

Limitations of 3D Scanners
Unfortunately, 3D scanners are not so versatile that they can be used on all types of surfaces and in all conditions. Like any piece of equipment, there are applications for which they may not be well suited. In some instances, measurements simply cannot be taken while in others, variations in measurements become visible to the eye such that technicians refer to these inaccuracies as "noise". Below are a list of items that need to be considered when working with 3D scan data for crime scene analysis and reconstruction:

Level of Required Accuracy. Every instrument has a range where it provides the most accurate results. Measurements taken beyond the practical range in given conditions may not be accurate and the results will be subject to scrutiny. One would not setup a 3D Scanner 50 meters away from a footprint that requires detailed analysis and study. It would clearly be better to take a separate scan of the footprint using a close range scanner less than one meter away.

Reflective Surfaces. Since most scanners are looking for a reflected beam of laser light, highly reflective and glossy surfaces cause a particular problem, redirecting light away from the 3D scanner. In some cases, it is possible for the beam of light to bounce off another surface and back to the scanner, giving erroneous results. Therefore, glass, water, and mirror-like objects usually case "noise" or will simply not provide any data.

Dark/Highly Absorptive Surfaces. Dark surfaces absorb light and make it difficult to get a strong reading to the receiver. In the worst case, a dark shiny surface (such as a clean, shiny black car) may be very difficult to scan without the surface of the car being treated. In many cases, scanning technicians will apply special waxes or white powdered sprays to absorptive surfaces. The powders are normally light in color so as to reflect more of the light back to the scanner.

Image of a vehicle that has been brushed with a white powder in order to assist scanner measurements.

Oblique Surfaces. When an object's surface is perpendicular to the laser scanner's receiver, it usually provides a strong signal back that can be accurately measured. However, as the surface of the object becomes more oblique, less of the beam is directed back to the laser scanner. Imagine a basketball being scanned. As the laser moves away from the center of the surface towards the top or bottom of the ball, the surface begins to move closer to a parallel surface (not perpendicular).

scanning Oblique Surfaces

Registration Errors. When combining several scans in one large point cloud, registration errors need to be considered because the final results may contain larger than expected errors. This can be noticeable when there are overlapping scans and certain features are clearly in two nearby places at once, but not so close as to have the points overlap exactly. The best solution is to use targets and reference markers during scanning as many scanners have the capability to align individual scans during the actual scanning and not while processing the data.

Conclusion
The future of 3D scanning technologies for the documentation and analysis of crime and accident reconstruction will undoubtedly continue to flourish. Most people think of 3D scanners as tools for larger objects such as roadways and buildings, however it is important to realize that scanners come in different sizes and measuring ranges from micrometers to several kilometers.

The impact of scanning technology in forensics is that there is now a migration towards photorealistic 3D environments where millions of points can be measured and analyzed. This is a slight departure from the traditional 2D photographs, sketches, and total station measurements of past. Yet, 3D scanning is still relatively new and the cost of a mid-range Lidar scanner can be expected to be more than $130,000 once the necessary hardware and software is included to make a functional field kit. This makes it rather difficult for some smaller agencies and private companies to invest in such equipment. However, this should not divert from other capable technologies such as photogrammetry, where the cost of a digital camera, software, and training may be no more than $5,000.

As manufacturers become more competitive in their offerings, the cost of 3D scanners will decrease, bringing more value and access to the field of forensics. Yet, the greatest area of development will be in the available software tools for the analysis and visualization of scan data. This is already becoming apparent in the market by some manufacturers and third party software developers who are providing specific packages aimed at assisting forensic technicians with customized analysis tools. Some of these tools apply to bullet trajectories, bloodstain pattern analysis, and even "image projection" where a suspect's height can be estimated from security videos. Once the analysis is complete, the results can be immediately displayed as a 3D animation or through an interactive viewer for court presentations. Eventually, the benefits of 3D scanners and related analysis tools will become more familiar to law enforcement agencies, attorneys, and jurors.

References:

1. “Laser Safety.” 30 Nov. 2009, Wikipedia. 3 Dec. 2009. http://en.wikipedia.org/wiki/Laser_safety#Class_1

Eugene Liscio, P. Eng, is the owner of AI2-3D. A company that specializes in 3D Forensic Measurement and Visualizations. He is a forensic animator and photogrammetry specialist. AI2-3D, Toronto, Ontario, 416-704-2695, eliscio@ai2-3d.com, www.ai2-3d.com.

Read Part 1 of this article: A Primer on 3D Scanning in Forensics: Part 1

Advertisement
Advertisement